CHARACTERIZATION OF TITANIUM ALUMINIDE ALLOY COMPONENTS FABRICATED BY ADDITIVE MANUFACTURING USING ELECTRON BEAM MELTING

goodyearmiaowMechanics

Nov 18, 2013 (3 years and 8 months ago)

72 views

L. E. Murr,
1,2

S. M. Gaytan,
1,2

A. Ceylan,
1

E. Martinez,
1,2

J. L. Martinez,
1,2

D. H.
Hernandez,
1,2

B.I. Machado,
1,2

D.A. Ramirez,
1

F. Medina,
2

S. Collins
3

and R.
B. Wicker
2,4


CHARACTERIZATION OF TITANIUM ALUMINIDE ALLOY
COMPONENTS FABRICATED BY ADDITIVE
MANUFACTURING USING ELECTRON BEAM MELTING


1
Department of Metallurgical and Materials Engineering

The University of Texas at El Paso, El Paso, TX 79968 USA



2
W. M. Keck Center for 3D Innovation

The University of Texas at El Paso, El Paso, TX 79968 USA



3
Additive Manufacturing Processes, 4995 Paseo Montelena, Camarillo, CA 93012 USA



4
Department of Mechanical Engineering

The University of Texas at El Paso, El Paso, TX 79968 USA


Abstract


Intermetallic,

γ
-
TiAl,

equiaxed,

small
-
grain

(~
2
µm)

structures

with

lamellar

γ/α
2
-
Ti
3
Al

colonies

with

average

spacing

of

0
.
6

µm

have

been

fabricated

by

additive

manufacturing

(AM)

using

electron

beam

melting

(EBM)

of

precursor,

atomized

powder
.

The

residual

microindentation

(Vickers)

hardness

(HV)

averaged

4
.
1

GPa

corresponding

to

a

nominal

yield

strength

of

~
1
.
4

GPa

(~HV/
3
),

and

a

specific

yield

strength

of

0
.
37

GPa

cm
3
/g

(for

a

density

of

3
.
76

g/cm
3
)

in

contrast

to

0
.
27

GPa

cm
3
/g

for

EBM
-
fabricated

Ti
-
6
Al
-
4
V

components
.

These

results

demonstrate

the

potential

to

fabricate

near

net

shape

and

complex

titanium

aluminide

products

directly

using

EBM

technology

in

important

aerospace

and

automotive

applications
.


Introduction


For at least the past decade, considerable efforts have been
made world
-
wide in the development, technology, and
applications of intermetallic, γ
-
TiAl based alloys: Ti
-

(34
-
49) Al


(5
-
10) Nb


(2
-
5) Cr (in atomic percent, a/o), at
high
-
temperatures for long term operation
[1
-
4]
.
Applications include propulsion exhaust system
components such as nozzle and divergent flaps fabricated
from wrought and cast gamma substructures
[5]
, and
aeroengine compressor blades. Other aerospace and
automotive applications have also been examined in recent
years
[4,6,7]

especially in the context of replacing the heavier,
nickel
-
based superalloys for the next generation of aircraft
engines, space vehicles, and automotive engine
components, including turbine wheels and engine exhaust
valves and pistons for improved auto fuel economy
[5
-
10]
.

Experimental and Analytical Issues


In this investigation we utilized an atormized, rapidly
solidified γ
-
TiAl
-
based


alloy powder with a nominal composition of Ti
-
47a/oAl +
2a/o Nb + 2a/o Cr (atomic percent: a/o)
17
. In contrast to
the single
-
phase, stoichiometric γ
-
Ti
-
55a/oAl eutectic, this
alloy was slightly lean in Al which creates a 2
-
phase
structure: γ
-
TiAl with ideally the tetragonal (L1
o
) (p4/mmm)
structure and the α
2
-
Ti
3
Al hexagonal (D0
19
) (p63/mmc)
structure (a = 0.58nm b = 0.46nm). With somewhat lower
Al content as in this study, this alloy produces a decreased
tetragonality to the point where c/a


1, and the γ
-
TiAl
structure is essentially ordered fcc (a


0.41 nm), with
alternating planes of Ti and Al in the [001] direction
[19,20]
.



Fig. 1. Precursor
titanium aluminide alloy
powder (a) and powder
particle size analysis (b).
The average powder
particle diameter is
indicated by the arrow
to be ~13 µm.


Fig. 2.
EBM
system
schematic.
Numbered
components and
functions are
discussed in the
text.



Sections cut from the EBM
-
produced test monoliths were also
ground and polished to a thickness of ~200 µm and 3 mm
diameter discs punched from these thinned samples which were
electropolished to electron transparency in a Tenupol 3
-
dual jet
electropolishing unit operated at a polishing current of 5
-
10A at
a voltage of ~25 V, using a polishing solution consisting of 900
mL ethanol and 50 mL HF at
-
20˚C. The electropolished discs
were examined in both a Hitachi H
-
8000 TEM or a Hitachi H
-
9500 TEM, operating at 200 or 300 kV, respectively, using a
goniometer tilt stage. The H
-
9500 TEM was fitted with digital
imaging camera and an EDAX
-
EDS system.



The residual hardness for the EBM
-
built test blocks was
measured using a Rockwell C
-
scale indenter (at 150 kg) as well as
a Vickers microindentation hardness tester (Struers Doramin
-
A300) using a l00 gf load, with a dwell time of 10s.


Results and Discussion


Figure 3 shows a typical EBM
-
built test
-
block
specimen. Figure 3(a) shows the build direction
(arrow) and the top surface while Fig. 3(b) shows a
magnified section showing vertical surface
particles (at S
V
) and the final horizontal surface
melt
-
scan features (S
n
). The measured density (ρ)
for the test blocks (measured weight/volume) as in
Fig. 3 was ~3.76 g/cm
3
. This compares favorably
with the ideal (or theoretical) density for this
composition of 3.84 g/cm
3
.


Fig. 3. Optical (a) and SEM (b) views of a typical test component. In
(a) the build direction is shown by the arrow. The melt surface at the
top of the build shown in (b) highlights the beam (melt) scan.


Fig. 4.
SEM views of the side edge of the test component in
Fig. 3(a) showing unmelted surface particles (a) and a
magnified view in (b).

4


Fig. 5. SEM views
looking down the
edge section for the
test component in
Fig. 3. (a) View of
horizontal (top
melt) and vertical
surfaces. (b)
Magnified view of
(a) showing
sintered and
partially melted
zone
corresponding to
the reference arrow
in (a).


Fig. 6. SEM views
of a polished and
unetched

horizontal section
from a test block
as in Fig. 3(a)
showing
unconsolidated
and
unmelted

regions creating
porosity in (a)
and (b) and a
remnant
Ar

bubble section at
arrow in (b)


Figure 7 shows optical metallographic views of the
microstructure typical of polished and etched
horizontal (Fig. 7(a)) and vertical (Fig. 7(b)) sections
from Fig. 3. The coarse lamellar γ/

2

colony structure
is also implicit from the comparative XRD spectra
shown in Fig. 8 for both the precursor powder (Fig.
8(a)) and the powder prepared (by fine filing) from the
EBM
-
fabricated test specimens (Fig. 8(b)) shown
typically in Fig. 3. It can be observed on comparing Fig.
8(a) and (b) that the initial (precursor) titanium
aluminide powder has a ratio γ/α
2


0.5 based on
approximate spectral (diffraction) peak heights
indicating a prominence of α
2
.


Fig. 7.
Optical
metallographic section
views in the horizontal
surface plane (a) and
the vertical surface
plane (b). The dark,
lamellar colonies are α
2
.


Fig. 8. XRD spectra
for the initial
(precursor) titanium
aluminide alloy
powder (a) and the
EBM
-
fabricated test
specimens (b).


The
equiaxed

grain structures in Fig. 7 have an average
grain size of ~2 µm in both the vertical and horizontal
directions, and there is correspondingly no extended,
columnar growth phenomena in the build direction
(vertical direction in Fig. 2(a)). There is also no clear melt
layer delineation. Correspondingly, and as noted above,
Fig. 9 shows that the chemistry of the precursor powder
(Fig. 9(a)) and the internal (vertical: Fig 9(b)) and
horizontal (surface melt: Fig. 9(c)) plane for the test
specimen shown in Fig. 3 is essentially unchanged. This is
in contrast to the earlier work by Cormier, et al.
[17]

where
the atomic weight aluminum decreased from 46 to 39
percent (~15%) on comparing the precursor powder
chemistry with the EBM fabricated test specimen.


Fig. 9. Energy
-
dispersive
X
-
ray spectra for the
initial (precursor)
titanium aluminide alloy
powder (a) and the
interior (b) and surface
(c) for the EBM
-
fabricated test specimens
in Fig. 3. The peaks
specify the atomic
percent of each
represented element in
parentheses.



Figure 10 shows TEM (300 kV) bright
-
field images for thin
films prepared from a horizontal (plane) slice from the test
specimen in Fig. 3, corresponding to the optical
metallographic view in Fig. 7(a). Figure 10(a) shows
primarily fine
-
grained and deformed γ (as evidenced by a
relatively high dislocation density) while Fig. 10(b) shows

2
-
Ti
3
Al laths in the γ
-
TiAl matrix. Figure 11 illustrates
deformation micro
-
twins and a relatively heavy dislocation
density (~10
9
/cm
2
) in a γ grain. Similar deformation
behavior was observed in (

2

+ γ) titanium aluminides by
Appel, et al.
[2]
. This deformation behavior is a reflection of
the rapid solidification phenomena associated with the
EBM
-
layer building process and accounts for the measured
hardness for the EBM fabricated test specimens (Fig. 3).


Fig. 10.
TEM bright
-
field
images for EBM
-
fabricated titanium
aluminide alloy
specimens. (a) γ
-
TiAl
grain and subgrain
microstructures,
including dislocation
substructures. (b)

2
-
Ti
3
Al lamellar laths in γ
-
phase grain structure.
(300kV accelerating
voltage).



Fig. 11.
TEM bright
-
field image of thin
deformation (micro)
twins in a γ
-
TiAl
grain with heavy
dislocation
substructure. (300kV
accelerating voltage)

Summary and Conclusions


In this study we have characterized the structures and
microstructures for precursor titanium aluminide powder
and solid 2
-
phase titanium aluminide components
fabricated by EBM. The precursor powder was α
2
-
phase
-
rich (hcp Ti
3
Al) while the EBM fabricated test components
were largely γ
-
TiAl (fcc). The EBM fabricated test
specimens exhibited an equiaxed γ
-
TiAl grain structure
with lamellar γ/α
2

colony structure having an average
spacing of 0.6 µm within the γ
-
grains which had an average
size of ~2 µm. However, TEM analysis illustrated α
2

laths
spaced from ~0.1 µm to 0.3 µm in addition to thin,
deformation microtwins in the γ
-
grains. A relatively high
dislocation density also contributed to the residual
hardness, and was due in part to the rapid cooling
associated with the EBM
-
layer fabrication process.



While the EBM
-
fabricated titanium aluminide
prototype density was ~3.76 g/m
3

or roughly 98% of
theoretical density, there were some regions of
residual porosity due to slightly non
-
optimized build
parameters as well as residual Ar bubbles with
maximum diameters of roughly 40 µm. These EBM
-
fabricated titanium aluminide prototypes had an
average microindentation hardness of 4.1 GPa,
corresponding to a yield stress of ~1.4 GPa and a
specific yield strength of ~0.37 GPa cm
3
/g, in contrast
to ~0.27 GPa cm
3
/g for Ti
-
6Al
-
4V EBM
-
fabricated
components.


[1]


Kim Y
-
W, Dimiduk DM, Loretto MH (Eds.), Gamma Titanium

Aluminides 1999, TMS, Warrendale, PA 1999.

[2]


Nathal MV, Darolia R, Liu CT, Martin PL, Miracle DB, Wagner R,

Yamaguchi M, (Eds.), Structural Intermetallics, 1997, TMS, Warrendale,

PA 1997.

[3]


Appel F, Brossman U, Christoph U, Eggert S, Janschek P, Lorenz U,

Müllauer J, Ochring M, Paul JDH, Adv. Eng. Mater. 2000; 2(1):699.

[4]

Loria EA, Intermetallics 2001, 9(2): 997.

[5]


Bartolutta PA, Krause DL, Titanium Aluminide Applications in the High

Speed Civil Transport, NASA/TM
-
1999
-
209071, NASA
-
Glen Research

Center, March

1999. [6]

Clemens H, Kesstler H, Adv. Eng. Mater.

2000; 2(9):551.

[7]


Kim Y
-
W, Clemens H, Rosenberger AH (Eds.), Gamma Titanium

Aluminides 2003, TMS, Warrendale, PA 2003.

[8]

Tetsui T, in Gamma Titanium Aluminides 1999 (Eds: Kim Y
-
W, Dimiduk

DM, Lorretto MH, TMS, Warrendale, PA 1999, p. 15.

9]


Noda T, Intermetallics 1998; 6:709.


[10]

Bauer H, Joos R, Smarsly W, Clemens H, in Intermetallics and

Superalloys (Eds:

Morris DG, Naka S, Caron P), WI
-
LEY
-
VCH,

Weinheim 2000, p. 384.

[11]

Appel F, Ochring M, Paul JDH, Klinkenberg Ch, Carneiro T,

Intermetallics 2004; 12(7
-
9):798.

[12]

Güther V, Otto A, Kestler H, Clemens H, in Gamma Titanium

Aluminides 1999

(Eds: Kim Y
-
W, Dimiduk DM, Loretto MH) TMS,

Warrendale, PA 1999, p. 225.

[13]

Clemens H, Kestler H, Eberhardt N, Knabl W, ibid, p. 209.

[14]

Clemens H, Metallkd. Z, 1995; 86:814.

[15]

Larsen DE, Wheeler DA, London B, in Proceedings of Processing and

Fabrication of Advanced Materials III, TMS, Warrendale, PA 1994,

p. 631.

[16]

Gerling R, Schimansky F
-
P, Wagner R, in Advances in Powder

Metallurgy & Particulate Materials (Eds: Capus JM, German RM),

MPIF, Princeton, NJ 1992,

p. 215.

[17]

Cormier D, Harryson O, Mahale T, West H, Res. Lett. Mater. Sci. 2007;

2007:1.

[18]

Liu W, DuPont J, in Rapid Prototyping of Materials (Eds: Marquis FDS,

Bourell DL) TMS, Warrendale, OH, 2002, p. 95.

[19]

Inui H, Nakamsra A, Oh MH, Yamaguchi M, Ultramicroscopy 1991;

39:268.

[20]

Gaytan SM, Murr LE, Medina F, Martinez E, Lopez MI, Wicker RB,

Mater.

Technol.: Adv. Perform. Mater. 2009, in press.

[21]

Murr LE, Gaytan SM, Medina F, Martinez E, Hernandez DH, Martinez L,

Lopez MI, Wicker RB and Collins S in Proceedings of 20
th

Int. Solid

Freeform Fab. Symposium, Austin, TX 2009.

[22]

Kestler H, Clemens H, Baur H, Joos R, Gerling R, Cam G, Bartels A,

Schleinzer C, Smarsly W, in Gamma Titanium Aluminides 1999 (Eds:

Kim Y
-
W, Dimiduk DM, Loretto MH), TMS, Warrendale, PA 1999, p.

423.

[23]

Murr LE, Interfacial Phenomena in Metals and Alloys, Addision
-
Wesley

Publishers, Reading, MA. 1975, p. 285.

[24]


Choi RW, Deng YG, McCullough C, Paden B and Mehrabian R, Acta

Metall.



Mater. 1990; 38(11): 2225.

[25]

Moll JH, Whitney E, Yolton C and Habel U, in Gamma Titanium

Aluminides 1999

(Eds. Kim, Y
-
W, Dimiduk DM, Loretto MH), TMS,

Warrendale, PA 1999, p. 255.

[26]

Srivastava D, Chang ITH and Loretto MH, Intermetallics 2001;

179(12):1003.

[27]

Murr LE, Gaytan SM, Medina F, Lopez H, Martinez E, Machado BI,

Hernandez DH,

Martinez L, Lopez MI, Wicker RB, Bracke J, Phil. Trans.

Roy. Soc. (London) 2009; A: in press.